Introduction to Applied Machine Learning

Machine learning here is framed as a decision problem: given some data and a task, figure out which family of algorithm fits, train it sensibly, and prove it actually works. You'll work through weekly homeworks on the standard supervised and unsupervised methods (regression, trees, SVMs, KNN, neural nets, clustering), then pull it together in a term project where you build and present an end-to-end pipeline on a dataset of your choice. The course sits at the applied end of CTIS, so the emphasis is less on deriving the math and more on picking the right method, choosing honest performance metrics, and reading results without fooling yourself.